What system changes would be necessary to make spellcasting checks workable?

[TuggyNE]

So, Truenaming, along with various generally-ill-fated homebrew attempts to balance arcane/divine casting, tried to base casting off the results of a skill check. It doesn't work well in 3.x because skill checks have a very low optimization floor and a very high optimization ceiling, so characters using the subsystem tend to either win at everything the subsystem can do, lose hard, or carefully balance on the ragged edge of "I'm trying really hard not to break the game, OK?".

However, the idea is tempting for a number of reasons: it's more elegant than ASF or universal SR, more even-handed than giving buffs a free pass on all points of failure (saves/SR/attack rolls), more interesting than fire-and-forget spell slots, more flexible than infinite uses of two or three limited-scope abilities, and may lead to the Philosopher's Stone of balancing casters.

So what changes would need to be made to make some sort of d20 check work out? Skill checks have magic items, class features, and spells that throw off the RNG by up to 100 points or more; ability checks don't scale well with level, and while more limited, still have a few ways to cheese them off the RNG; CL checks can be pumped almost as easily as skills, and seem thematically off (given how CL is often used for the spell's effect in some way as it is). Ideally, the check should allow for specialization in subsets of magic (such as different skills for Pyromancy than for Scrying, or something) and should have approximately the same variability in effectiveness of melee attack rolls (without touch attacks or Shock Trooper, anyway).

And before you suggest it, relying solely on action economy to keep (spellcasting check-boosting) buffs in line is something of a losing battle.

[Jack_Simth]

Well, the first step would be to get rid of anything that could boost the skill check. No, you can't make a custom item of +10 Casting Skill. They don't work. Things that normally boost all skills (such as Heroism and it's Greater cousin) or that boost an arbitrary skill (such as Divine Insight) don't work for Casting Skill because of *Oh look, an obvious distraction*
[Runs away.]

[eggynack]

Ideally, the check should allow for specialization in subsets of magic (such as different skills for Pyromancy than for Scrying, or something) and should have approximately the same variability in effectiveness of melee attack rolls (without touch attacks or Shock Trooper, anyway).

Why not just do this? Instead of having one big, "Put all the optimization here," skill, why not have several? I don't have any specifics, but you could split it up so that you couldn't just optimize one check to get all the broken things. That way, you could set the checks up so that they're easier to reach than truenamer checks, but so that it's really expensive to optimize them all simultaneously. You could even key them off of something like knowledge, or if that's too easy to optimize in a generic way, just have different versions of spellcraft. In fact, you could set it up so that learning spells would require the same spellcraft types as casting them, so that things would fit together..

[Jeff the Green]

Why not just do this? Instead of having one big, "Put all the optimization here," skill, why not have several? I don't have any specifics, but you could split it up so that you couldn't just optimize one check to get all the broken things. That way, you could set the checks up so that they're easier to reach than truenamer checks, but so that it's really expensive to optimize them all simultaneously. You could even key them off of something like knowledge, or if that's too easy to optimize in a generic way, just have different versions of spellcraft. In fact, you could set it up so that learning spells would require the same spellcraft types as casting them, so that things would fit together..

The problem is that, even if you get 2 Skillpoints per level, any caster worth their salt is going to start with at least 18 Intelligence. Assuming you also need Concentration, that means you still have 5 skills to max out.

Also, ranks are probably the least important part of optimization. The big problem is that skill bonus items cost Bonus squared x 100 GP. So a +23 (equivalent to max ranks at level 20) item only costs 52,900 GP. And then there's guidance of the avatar, divine inspiration, heroism, etc.

Honestly, I think in order for this to work, you'd have to revamp most of the skill system.

[Roguenewb]

I use this with Spellcraft in some campaigns.

To cast a spell, a caster must make a DC 15+2 x spell level.

Special: No bonuses from sources other than Class Features, Feats and Intelligence modifier apply to this skill.

Special: If a class uses an ability other than intelligence to determine spell save DC, they may add that modifier as an additional bonus to this check.

Hurts wizards most. They lose some efficacy from actually losing actions in combat. Failing the check doesn't lose the spell, they just lose the action. If they op INT and get Spell Focus Spellcraft they have like a 15% chance of losing a given spell that decreases as they magic up their int.

[eggynack]

The problem is that, even if you get 2 Skillpoints per level, any caster worth their salt is going to start with at least 18 Intelligence. Assuming you also need Concentration, that means you still have 5 skills to max out.

Also, ranks are probably the least important part of optimization. The big problem is that skill bonus items cost Bonus squared x 100 GP. So a +23 (equivalent to max ranks at level 20) item only costs 52,900 GP. And then there's guidance of the avatar, divine inspiration, heroism, etc.

Honestly, I think in order for this to work, you'd have to revamp most of the skill system.
I think that the trick here is that you require a certain amount of actual optimization to hit the skill checks. You only need that 52,900 GP to optimize one skill, but if it's eight skills instead, that's a lot more difficult. I don't know where the balance point is, between trivial and truenamer, but I think it exists. See, it's easy to just win against one skill check, but it might be harder to win against a pile. If you really wanted to go all out, you could key the skills against different stats or something. I don't know how far you can go without going over the top, but ideally a wizard should just about always hit the checks within his specialization, hit the checks most of the time in his second specialization, and then it should go down to around 50%, before dipping down towards zero. Additionally, this would require developing new spell schools, so that the 100% isn't attached to conjuration, with the 80% attached to transmutation. It'd be complicated to put together, but it seems doable.

[TuggyNE]

Special: No bonuses from sources other than Class Features, Feats and Intelligence modifier apply to this skill.

This is the part that feels off to me, since the point of having a unified system is to reduce weird special cases like this. After all, why are you calling it a "skill check" if the various things that interact with "skill checks" don't work on this? (What's more, "class features" include known cheese sources like Motivate Intelligence or, less horribly, Inspire Competence.) Instead, as Jeff the Green suggests, revamping the skill system so no skills are stupidly pumpable like this would probably work a lot better.


I think that the trick here is that you require a certain amount of actual optimization to hit the skill checks. You only need that 52,900 GP to optimize one skill, but if it's eight skills instead, that's a lot more difficult. I don't know where the balance point is, between trivial and truenamer, but I think it exists. See, it's easy to just win against one skill check, but it might be harder to win against a pile. If you really wanted to go all out, you could key the skills against different stats or something.

It's not all that much more difficult to get that many more skill boosters; you might even be able to shove them all together on the same magic item without a cost increase, depending on guideline interpretation. In any case, WBL and especially crafting makes it quite practical to optimize lots of skills simultaneously.


I don't know how far you can go without going over the top, but ideally a wizard should just about always hit the checks within his specialization, hit the checks most of the time in his second specialization, and then it should go down to around 50%, before dipping down towards zero. Additionally, this would require developing new spell schools, so that the 100% isn't attached to conjuration, with the 80% attached to transmutation. It'd be complicated to put together, but it seems doable.

Hmm, yes. I'm thinking of scaling DCs at something vaguely like 15 + triple spell level, which puts it a bit below Truenamer's rate and links it to spell power rather than enemy danger. That's only workable if skill pumping is stripped of its silly excesses, obviously.

So I guess this is kind of three related projects, really: fix (skill/ability/CL) boosting, rewrite spell schools, and finally impose a check for successful casting.

[Amidus Drexel]

Additionally, this would require developing new spell schools, so that the 100% isn't attached to conjuration, with the 80% attached to transmutation. It'd be complicated to put together, but it seems doable.

Do it by subschool, then, and have a [general] tag for each regular school to cover the spells without any. You have a Casting [teleportation] skill, a Casting [summoning] skill, a Casting [creation] skill, a Casting [healing] skill, and a Casting [calling] skill, for example.

I think it would even up some of the differences between schools, too, as ones with less subschools (like say, abjuration), will be less affected than the schools with a wide variety of subschools (like conjuration).

[Jeff the Green]

Hmm, yes. I'm thinking of scaling DCs at something vaguely like 15 + triple spell level, which puts it a bit below Truenamer's rate and links it to spell power rather than enemy danger. That's only workable if skill pumping is stripped of its silly excesses, obviously.

So I guess this is kind of three related projects, really: fix (skill/ability/CL) boosting, rewrite spell schools, and finally impose a check for successful casting.

Another idea would be to make a difference between combat and utility. I feel like it should be trivial to cast Mass CLW to heal up after battle, but significantly more difficult to revivify in battle.

Also, this would make Exemplar a mandatory dip for spellcasters and make Sorcerers even more gimped compared to Wizard than they already are.

[Roguenewb]

This is the part that feels off to me, since the point of having a unified system is to reduce weird special cases like this. After all, why are you calling it a "skill check" if the various things that interact with "skill checks" don't work on this? (What's more, "class features" include known cheese sources like Motivate Intelligence or, less horribly, Inspire Competence.) Instead, as Jeff the Green suggests, revamping the skill system so no skills are stupidly pumpable like this would probably work a lot better.



It's not all that much more difficult to get that many more skill boosters; you might even be able to shove them all together on the same magic item without a cost increase, depending on guideline interpretation. In any case, WBL and especially crafting makes it quite practical to optimize lots of skills simultaneously.


It has to be there. It is simply too easy for skill bonuses to be pumped into the stratosphere. Class features like inspire competence or motivate intellegence almost always require investment by the player, lowering their overall power level, a good thing for tier 1s and 2s. I guess disallowing custom skill items may help a fair amount, but it isn't perfect.

[gorfnab]

I use this with Spellcraft in some campaigns.

To cast a spell, a caster must make a DC 15+2 x spell level.

Special: No bonuses from sources other than Class Features, Feats and Intelligence modifier apply to this skill.

Special: If a class uses an ability other than intelligence to determine spell save DC, they may add that modifier as an additional bonus to this check.

A variant I've kicked around a few times based on a version from one of the DMs I played at college:

No Spell Slots. As long as you can make the Spellcraft check you can cast the spell. Abilities that use a Spell Slot are treated as if you were trying to cast a spell of the required level.

Spellcraft DC to cast a spell = 10 + level of spell you are trying to cast + number of spells you have cast that day (does not include cantrips, see below).

Spellcraft DC to cast a cantrip (0th level spell) = 10 + number of cantrips that you have cast that day.

Spellcraft skill is based on the spell save DC stat of the class. Magic items do not confer bonuses to this check. Only ability score, feats, and class features may modify the Spellcraft skill.

Spells known gained for full spellcasters (and Bards) at casting mod (unmodified by magic items, minimum 2) per level. Cantrips (0th level spells) known is equal to casting mod (unmodified by magic items, minimum 2) at first level with a bonus cantrip known on levels that are multiples of 4 (4, 8, 12, etc...). So a first level Cleric with an 18 Wisdom would know 4 cantrips and 4 first level spells. At 2nd level that Cleric would add 4 spells of any level they can cast to their spells known list (so 4 more first level spells, or 2 first level spells and 2 cantrips, etc...). For full spellcasters follow the chart for Cleric to determine level of spells you can cast (Bards follow the Bard chart). Half casters such as ranger gain one spell known every level starting at the level they can cast spells.

Spellcasters do not have to prepare spells. You can cast any spell you know if you can make a successful Spellcraft check. Spellcasters regain spells (basically reset spells cast that day) after 8 hours of rest and 15 minutes of meditation or prayer (similar to a sorcerer). Any class that can cast spells but does not have Spellcraft as a skill gains Spellcraft as a class skill.

New class: Mage (gestalt of Sorcerer/Wizard) - basically Saves, BAB, HD, Familiar, Skill points per level, and Proficiencies of a Sorcerer. Skill list is the combined list of Sorcerer and Wizard skills. Scribe Scroll and bonus feats of a Wizard. Spellcasting stat can be Int or Cha chosen at first level. Ability to switch out spells known on the even levels as a Sorcerer.

[TuggyNE]

It has to be there. It is simply too easy for skill bonuses to be pumped into the stratosphere. Class features like inspire competence or motivate intellegence almost always require investment by the player, lowering their overall power level, a good thing for tier 1s and 2s. I guess disallowing custom skill items may help a fair amount, but it isn't perfect.

Well, let's put it this way: there needs to be some way of preventing the absurd skill-boosting, but putting in a hack to the effect that "this skill, unlike every other skill, somehow magically can't be affected by stuff that's specifically made to improve all skills, because, uh, reasons" seems frankly rather lame. Instead, I'd prefer something that actually fixes the problem: skill-boosting shenanigans. (Note, too, that while Inspire Competence etc do require investment, that investment need not, and in some cases cannot, be made by the actual caster in question. Consider a party of a Wizard, a Cleric, a Warmage, and a Bard/Marshal; the Bard can slap Inspire Competence on whichever caster needs it most, and Motivate Intelligence for everyone. That's a perfectly plausible setup that requires no special effort from anyone but the Bard, and not much even there. And it's putting +5 to +9 or more on checks, which is pretty substantial.)


(assorted)

That seems interesting, although again it suffers from the arbitrary, artificial restriction on one skill. Is there a comparison writeup of likely spells/day with the system versus Vancian?

I do note one apparent problem: it's best to cast the highest-level spells you can as early and often as possible, since there is no difference between successfully casting a high-level spell and a low-level spell: both add +1 to future DCs, and no more. I'm not sure what, if anything, can be done about that unfortunate quirk.

[eggynack]

Well, let's put it this way: there needs to be some way of preventing the absurd skill-boosting, but putting in a hack to the effect that "this skill, unlike every other skill, somehow magically can't be affected by stuff that's specifically made to improve all skills, because, uh, reasons" seems frankly rather lame.

What if the (possibly arbitrary) restriction weren't that you can't get nifty skill boosting items at all? Instead, there would be some form of limit on how many different skill boosting capabilities you could stack on a single item. That way, it would be progressively harder to boost skill checks, the further afield of your chosen discipline you are. That'd be kinda interesting.

[Jeff the Green]

What if the (possibly arbitrary) restriction weren't that you can't get nifty skill boosting items at all? Instead, there would be some form of limit on how many different skill boosting capabilities you could stack on a single item. That way, it would be progressively harder to boost skill checks, the further afield of your chosen discipline you are. That'd be kinda interesting.

The problem is that there are already items that exist for nearly every other skill, if not indeed every other skill. You'd have to explain why such items don't exist for this skill/these skills in particular.

[undead hero]

A variant I've kicked around a few times based on a version from one of the DMs I played at college:

No Spell Slots. As long as you can make the Spellcraft check you can cast the spell. Abilities that use a Spell Slot are treated as if you were trying to cast a spell of the required level.

Spellcraft DC to cast a spell = 10 + level of spell you are trying to cast + number of spells you have cast that day (does not include cantrips, see below).

Spellcraft DC to cast a cantrip (0th level spell) = 10 + number of cantrips that you have cast that day.

Spellcraft skill is based on the spell save DC stat of the class. Magic items do not confer bonuses to this check. Only ability score, feats, and class features may modify the Spellcraft skill.

Spells known gained for full spellcasters (and Bards) at casting mod (unmodified by magic items, minimum 2) per level. Cantrips (0th level spells) known is equal to casting mod (unmodified by magic items, minimum 2) at first level with a bonus cantrip known on levels that are multiples of 4 (4, 8, 12, etc...). So a first level Cleric with an 18 Wisdom would know 4 cantrips and 4 first level spells. At 2nd level that Cleric would add 4 spells of any level they can cast to their spells known list (so 4 more first level spells, or 2 first level spells and 2 cantrips, etc...). For full spellcasters follow the chart for Cleric to determine level of spells you can cast (Bards follow the Bard chart). Half casters such as ranger gain one spell known every level starting at the level they can cast spells.

Spellcasters do not have to prepare spells. You can cast any spell you know if you can make a successful Spellcraft check. Spellcasters regain spells (basically reset spells cast that day) after 8 hours of rest and 15 minutes of meditation or prayer (similar to a sorcerer). Any class that can cast spells but does not have Spellcraft as a skill gains Spellcraft as a class skill.

New class: Mage (gestalt of Sorcerer/Wizard) - basically Saves, BAB, HD, Familiar, Skill points per level, and Proficiencies of a Sorcerer. Skill list is the combined list of Sorcerer and Wizard skills. Scribe Scroll and bonus feats of a Wizard. Spellcasting stat can be Int or Cha chosen at first level. Ability to switch out spells known on the even levels as a Sorcerer.

I love this, but I might adapt it to a caster level check or make a new skill based on con (concentration but you can't boost it via items and such).

Incantrix might also be the way to allow metamagic effects... Hmm.

[eggynack]

The problem is that there are already items that exist for nearly every other skill, if not indeed every other skill. You'd have to explain why such items don't exist for this skill/these skills in particular.
I'm not saying that the items don't exist. You can get an item of spellcraft (summoning) as easily as you would an item for any other skill. However, that same item can't also give a bonus to spellcraft (evocation). There would need to be an explanation for that, but it'd be a much less difficult one to sell. The net effect of this is that you'd need to stick your skill boosters in a number of different slots, and on some of those you'd need to pay a premium for an off slot, or no slot item. Furthermore, if you subdivide the required skills enough, it becomes nearly impossible to cover everything at once, maybe even at high levels. In fact, the point is that these items exist, so it becomes a tax on your ability to cast a variety of spell types.

[limejuicepowder]

In order to dodge the inherent and known problems of skill checks, it might be worth it to make a sub-set of "spell skills." Only caster classes would gain the ranks to spend on spell skills, and it would be separate from their pool of regular skills.

I would make several different spell skills, each representing a different aspect of magic. "Elemental," "Calling," "Creation," "Energy," "Senses," etc. Each spell would then have a DC check to be met with the associated skill. Complex spells might even require multiple checks based on different aspects.

Yeah this system would be a pain to implement.

[Fouredged Sword]

A slotless system of magic would be interesting for a high magic, high fantasy setting.

Consider the following skills
-
Abjuration
Conjuration
Divination
Enchantment
Evocation
Illusion
Necromancy
Transmutation

For each 2 skill points you invest in a skill you learn a new spell. That spell can be used any number of times per day. DC's ignore attributes, but go up by increasing the skill target DC 2 per 1 increase save dc. Spells have an activation DC of 6*spell level. You can learn a spell of level Rank / 4 spell level. There are no spells over 6th level.

In fact, one could bundle saves, bab, and other things into skills, strip the system down to one mechanic, and possibly even a single generic class with a number of skillpoints per level, and one feat per level.

Melee - You add this to your attack roll, each 3 ranks adds one to damage. At 6 ranks and each 5 ranks after you get an iterative
Ranged - As melee, but for ranged weapons
Will - ability to resist mental influence. roll this to resist mental effects
Fortitude - The ability to resit physical effects. Each 4 ranks invested in this grants 1 HP per HD.

It could make for an interesting game.

[Jeff the Green]

I'm not saying that the items don't exist. You can get an item of spellcraft (summoning) as easily as you would an item for any other skill. However, that same item can't also give a bonus to spellcraft (evocation). There would need to be an explanation for that, but it'd be a much less difficult one to sell. The net effect of this is that you'd need to stick your skill boosters in a number of different slots, and on some of those you'd need to pay a premium for an off slot, or no slot item. Furthermore, if you subdivide the required skills enough, it becomes nearly impossible to cover everything at once, maybe even at high levels. In fact, the point is that these items exist, so it becomes a tax on your ability to cast a variety of spell types.

That's more reasonable, I think. Maybe even make them have different bodyslot affinities. Face for Divination, hands for Illusion, torso for Polymorph and Necromancy, head for Enchantment, boots for Teleportation, belt for generic Transmutation and Conjuration, arms for Abjuration, throat for Summoning and Calling. Evocation could be slotless for free, probably.

One other problem I thought of. Are we really going to force Beguilers to buy an entire other item for glitterdust, Dread Necromancers one for dispel magic, or Warmages for whatever they pick up with eclectic learning?

[eggynack]

That's more reasonable, I think. Maybe even make them have different bodyslot affinities. Face for Divination, hands for Illusion, torso for Polymorph and Necromancy, head for Enchantment, boots for Teleportation, belt for generic Transmutation and Conjuration, arms for Abjuration, throat for Summoning and Calling. Evocation could be slotless for free, probably.
Maybe, yeah. The big problem with that is that you can just give these items abilities other than skill boosters, and bypass the issue entirely. Thus, you'd largely be free of the premium. That's why I figured that you'd mostly have them aligned to the same body slot, so that you'd be forced to pay premiums really fast. In fact, you could force the issue and have a character only be capable of using one of each overarching variety of skill boosting item, so that the wizard would suffer from diminishing returns as an inherent quality of the system.


One other problem I thought of. Are we really going to force Beguilers to buy an entire other item for glitterdust, Dread Necromancers one for dispel magic, or Warmages for whatever they pick up with eclectic learning?
I've gotta figure that tier 3 and 4 spontaneous casters could just do their stuff normally, without checks at all. It could also be neat to let sorcerers key all of their skill requirements off of bluff or something, so that they'd gain an advantage over wizards in that department. Sorcerers get to cast a small number out of a wide variety of spells, while wizards get to cast a massive number out of a narrow band of spells. If there's really a need for beguilers to have the skill check at all, they could also get the same limitation.

[TuggyNE]

Cursory reply because I'm sleepy but don't want to leave the thread hanging until tomorrow.


Furthermore, if you subdivide the required skills enough, it becomes nearly impossible to cover everything at once, maybe even at high levels. In fact, the point is that these items exist, so it becomes a tax on your ability to cast a variety of spell types.

It's not the worst idea, since it involves minimal changes to the system, but I can't say I'm terribly fond of skill item taxes either, especially given how hard it is to figure out exactly how much someone will spend. Consider someone spending 10000gp: they can buy a single +10 boost item in its proper slot and what-not, buy a single +8 item in the wrong slot or some such, buy four +5 items, or…. And that's at comparatively low levels, where having a difference of 5 or more is already a pretty significant swing, but in high levels you might be looking at variations between characters of +15, +20, +25.


In order to dodge the inherent and known problems of skill checks, it might be worth it to make a sub-set of "spell skills." Only caster classes would gain the ranks to spend on spell skills, and it would be separate from their pool of regular skills.

I would make several different spell skills, each representing a different aspect of magic. "Elemental," "Calling," "Creation," "Energy," "Senses," etc. Each spell would then have a DC check to be met with the associated skill. Complex spells might even require multiple checks based on different aspects.

Hmm, interesting. Ranks isn't really the problem, though; if you called it something that wasn't "skill", it might serve to segregate it properly. (Since there are a great deal fewer spells/class features/feats that give bonuses to checks that are neither attack rolls, saves, CL checks, ability checks, nor skill checks.)


It could make for an interesting game.

Consider me intrigued, although note that the tentative scaling is off; DCs scale by 1 per 2 modifier, while "saves" scale by 1 per 1 modifier, for example.


One other problem I thought of. Are we really going to force Beguilers to buy an entire other item for glitterdust, Dread Necromancers one for dispel magic, or Warmages for whatever they pick up with eclectic learning?

Yes! (No.)


I've gotta figure that tier 3 and 4 spontaneous casters could just do their stuff normally, without checks at all. It could also be neat to let sorcerers key all of their skill requirements off of bluff or something, so that they'd gain an advantage over wizards in that department. Sorcerers get to cast a small number out of a wide variety of spells, while wizards get to cast a massive number out of a narrow band of spells. If there's really a need for beguilers to have the skill check at all, they could also get the same limitation.

It seems inelegant to only partly replace Vancian casting. But eh.

[Fouredged Sword]

Tuggyne - You make a point about scaling, though in my mind, 3.5 has a problem with save or die. I think that spells should tilt the encounter to one side or the other without dominating the combat. This goes doubly so for spellcasters without spell slots. As spellcasters get access to stronger and stronger effects, they find that the targets of their spells have a much greater ability to resist the effects. Also, I intend for there to be enough skills that a character feels they need that they are forced to choose between saves, utility, and offense. Fully investing in all saves would cost nearly half your advancement each level, and would leave little else to do.

I could also see an implementation of exceeding the spell check DC adding to the save, rather than risking failure by increasing the DC by intent. Though that is getting to the point that maybe the system should just be a matter of opposed rolls for combat spells, with set DC's for buffs and utility spells. Maybe a modifier for higher or lower spells.

Also, spells shouldn't necessary be opposed by a save skill. Illusions should target the perception of the victim for example. No willsave or reflex or fortitude will allow your character to see through that invisibility.

Though I do like the idea of spells having enhancements you can apply to them by increasing the DC of the spell check.

[Grod_The_Giant]

So what changes would need to be made to make some sort of d20 check work out?
You'd have to really control skill bonuses. Like, make it impossible to boost beyond the expected bonuses-- otherwise numbers get out of control, and anything that does boost casting checks becomes required for all casters.

You also have to worry about requisite investment. The more distinct casting skills you offer, the more casters will have to pump Int to cover all their bases, and the less they can invest in non-casting abilities.

A better option, perhaps, is to make casting checks a class feature. Say you have 8 "skills*," for the 8 different schools. Each level, you get some number of "skill points*" which may be distributed as normal, and each skill has a cap based on your level.

A wizard, perhaps, might get 6 points per level, with a cap of level in all schools. A specialist might get a +1 bonus to one check and a -2 penalty to all the others. A sorcerer might get 4 points per level, with a cap of level +2 in one school. I don't know, I'm just making stuff up now, but you get the idea.

*Use different names to avoid confusion

[Deepbluediver]

There are two approaches to this, I think.

The first is to seperate out the check for casting from skills entirely, and make it its own statistic, like BAB. This is what I did for my magic overhaul.

Casting a spell requires a Wis check of either 10+twice spell level, or Spell level+SR, depending on if the spell targets a creature or not. (also, eveything has a base SR, just like base AC) You get a bonus to the check like BAB, except that the progression is inverted. Wizards have a bonus equal to their level, Fighters 1/2 their level, Clerics & Rogues 3/4 level, etc.

The link for the full magic-overhaul is in my extended sig if you want to peruse it for a more in-depth look of what I was talking about.


The second method, for a specific case where you wanted to link it to Skills (like the Truenamer), would be to spread out the checks to various skills, so as to divide the super-high levels of optimization. I suggested something like this to Grod in his "lots of little class tweaks" thread, but I think it was more complicated than what he was going for.

The spells (recitations? I can never remember what they're called) for a Truenamer that target creatures get a boost from the knowledge skills, depending on creature type.
Outsiders- Know. Religion
Constructs, Undead- Know. Arcana
Animals, Plants, Fey, Magic Beasts- Know (nature), etc.

The exception are the "nuturing word"s which are linked to Heal.

The landscape/perfected map ones can be boosted by Know. (local), Survival, Gather Information, possibly a few others.
The item-based ones are linked to various Craft skills, obviously.

Something like that might be a workable system, but probably only for a limited number of classes. I still prefer the first thing I outlined for a general spell-check.


Edit: After reading through prior responses, several other people have also explored the "split up the skills used" option, and noted that it still has potential issues. One thing about optimization is that it relies on having a magic-mart style campaign, so that the players can buy or build almost anything they want. In a setting where the DM keeps a tighter control over magic items, multiclassing, and splatbooks, the high-end of optimzation can be contained to some degree.

[Valwyn]

Probably not what you were asking for, but a system I know gives casters non-lethal damage each time they cast, making a Fort saves to resist the damage. If they pass the save by 10 or more, the damage is halved. If they pass the save by less than 10, they take normal damage. If they fail by less than 10, they take double damage and must pass a special check (1d20+CL+Stat vs the DC of the spell -10 (yes, spells have casting DCs)) or lose control of the spell (basically whatever the DM wants to happen: low-level spells fizzle, mid-level may daze/stun/knock down/other the caster, and high-level spells are not speficied, but most likely blow in your face or drag you to hell or whatever). If you fail by 10 or more, they take double damage, must pass the control check, and are fatigued (or exhausted if already fatigued).

Of course, the system has a few issues, like things immune to non-lethal damage and boosting your Fort saves to very high levels, but it might give you some ideas.

[TuggyNE]

Tuggyne - You make a point about scaling, though in my mind, 3.5 has a problem with save or die. I think that spells should tilt the encounter to one side or the other without dominating the combat. This goes doubly so for spellcasters without spell slots. As spellcasters get access to stronger and stronger effects, they find that the targets of their spells have a much greater ability to resist the effects.

I agree with the observation, but dropping caster DCs by 10 or more relative to saves over the course of levels is taking things much too far. To fix save-or-dies, I'd probably go with something like the will points in my homebrew sig, which reduces swinginess considerably. (I.e., the "wait random number of rounds before abruptly and utterly ending the fight" problem.)


Also, spells shouldn't necessary be opposed by a save skill. Illusions should target the perception of the victim for example. No willsave or reflex or fortitude will allow your character to see through that invisibility.

Hmm. Depends on the illusion, but yeah.


You'd have to really control skill bonuses. Like, make it impossible to boost beyond the expected bonuses-- otherwise numbers get out of control, and anything that does boost casting checks becomes required for all casters.

Not impossible, just difficult, like attack bonus is (as I noted earlier, in the absence of a few outliers like touch attacks). I don't really care if TO can double your effective spells per day or something, but if moderate optimization can do it that's a much more serious problem.


You also have to worry about requisite investment. The more distinct casting skills you offer, the more casters will have to pump Int to cover all their bases, and the less they can invest in non-casting abilities.

Some might consider that a feature, not a bug. :smallwink:


A better option, perhaps, is to make casting checks a class feature. Say you have 8 "skills*," for the 8 different schools. Each level, you get some number of "skill points*" which may be distributed as normal, and each skill has a cap based on your level.

A wizard, perhaps, might get 6 points per level, with a cap of level in all schools. A specialist might get a +1 bonus to one check and a -2 penalty to all the others. A sorcerer might get 4 points per level, with a cap of level +2 in one school. I don't know, I'm just making stuff up now, but you get the idea.

More or less, although by this point I'm thinking that rewriting the full-casting classes might be necessary, since the breaking out of skills along school lines means all casters need to have almost the same sort of school-awareness that Wizards do in Core.


*Use different names to avoid confusion

Resonance points, perhaps.


There are two approaches to this, I think.

The first is to seperate out the check for casting from skills entirely, and make it its own statistic, like BAB. This is what I did for my magic overhaul.

Casting a spell requires a Wis check of either 10+twice spell level, or Spell level+SR, depending on if the spell targets a creature or not. (also, eveything has a base SR, just like base AC)

I know I said giving everything SR was lame and inelegant, but upon thinking it over again, it might work — at least as long as it's unified with the idea that absolutely any spell has some sort of resistance to being cast. In other words, merely adding cursory SR to creatures fails because it's insufficient, but this idea sounds closer to what I'm thinking would work.


Edit: After reading through prior responses, several other people have also explored the "split up the skills used" option, and noted that it still has potential issues. One thing about optimization is that it relies on having a magic-mart style campaign, so that the players can buy or build almost anything they want. In a setting where the DM keeps a tighter control over magic items, multiclassing, and splatbooks, the high-end of optimzation can be contained to some degree.

I suppose that's true, but I'd rather not set up a system that only really works with low-op campaigns. After all, we already have one of those.


Of course, the system has a few issues, like things immune to non-lethal damage and boosting your Fort saves to very high levels, but it might give you some ideas.

Also, CL is boostable at mid-op and above, which throws off some of the checks, but yes. Again, though, this doesn't really curtail optimization; it merely reduces the effectiveness of those that aren't good enough at it. That's undesirable, since I don't care how strong low-op casters are, particularly.

[Grod_The_Giant]

Some might consider that a feature, not a bug. :smallwink:
Let me put that another way: The more you force casters to invest in specific areas to make their class features function, the more homogenous caster characters will be. Two sorcerers might be different due to different spell selection, but how different will two wizards be? Two druids?

[eggynack]

Let me put that another way: The more you force casters to invest in specific areas to make their class features function, the more homogenous caster characters will be. Two sorcerers might be different due to different spell selection, but how different will two wizards be? Two druids?
I think that you'd actually get less homogeneous casters. If you need to invest in order to access a broad spectrum of spells, and the spells in each set are roughly equal in power, then you're enforcing a pretty diverse variant on specialization, and the decisions based on when to expand into other schools, versus becoming better in one school, would be more interesting. You need to invest in specific areas to make class features function, but there would ideally be too many areas to invest in them all at once, so different wizards would have different class features.

[Amidus Drexel]

One of the earlier posts reminded me of a magic system that I was working on a couple months ago (I never finished it). The idea essentially made magic another subsystem available to every class, doing away with all of the full-casting classes.

To cast a spell, you needed ranks in both Spellcraft and one of Knowledge: Arcana, Religion, or Nature (depending on what kind of spell you wanted to cast; arcana for arcane spells, religion for cleric spells, and nature for druid spells), equal to the (spell's level+1)*2, giving a character with full investment in those skills 1st-level spells at level 1, 2nd-level spells at level 3, etc.

To cast a spell, you also needed an Int and Cha bonus equal to the spell level. (An Int&Cha score of 16 to cast 3rd level spells, for example). Getting 9th level spells is difficult (although not impossible) for humans, even with good starting scores.

Casting spells dealt Wis damage equal to the level of the spell cast. A 7th level spell, for example, would deal 7 Wis damage to the caster.

(I didn't have a way to model spells known, but using the sorcerer progression seems to make sense)

I can't say the system fixes spellcasting so much as drastically changes how it works, but I did like offering it to every class.

[TuggyNE]

To cast a spell, you needed ranks in both Spellcraft and one of Knowledge: Arcana, Religion, or Nature (depending on what kind of spell you wanted to cast; arcana for arcane spells, religion for cleric spells, and nature for druid spells), equal to the (spell's level+1)*2, giving a character with full investment in those skills 1st-level spells at level 1, 2nd-level spells at level 3, etc.

Tying it to ranks makes a fair bit of sense, since it's almost impossible to get extra ranks. (This being 3.x, almost nothing is impossible, of course.)


To cast a spell, you also needed an Int and Cha bonus equal to the spell level. (An Int&Cha score of 16 to cast 3rd level spells, for example). Getting 9th level spells is difficult (although not impossible) for humans, even with good starting scores.

Casting spells dealt Wis damage equal to the level of the spell cast. A 7th level spell, for example, would deal 7 Wis damage to the caster.

Wow, that's a tremendous cost. It doesn't seem like you'd want to cast spells unless there was basically no other option, or unless you cheesed your way out of the consequences somehow. I don't think I want to soft-ban casting, really.

[Amidus Drexel]

Tying it to ranks makes a fair bit of sense, since it's almost impossible to get extra ranks. (This being 3.x, almost nothing is impossible, of course.)

Wow, that's a tremendous cost. It doesn't seem like you'd want to cast spells unless there was basically no other option, or unless you cheesed your way out of the consequences somehow. I don't think I want to soft-ban casting, really.

True, although getting 9th-level spells at level 1 is possible with the spell slot system as it is. :smallamused:

It was meant to support low-level casting, while still allowing for higher-level spells to exist (at a significant cost - high-level magic is a powerful thing for people to be throwing around). It certainly doesn't support full-casting classes very well (which was my intention). I won't say it's perfect (far from it), but it does give everyone free cantrips with a few ranks in skills.

That, and repairing ability damage is fairly easy (depending on how much you have, of course) in 3.5. A dedicated caster with that system (probably a skill-focused character) would probably grab bind vestige and bind Naberius, or get an item that cast restoration a few times per day. *shrug* Again, not perfect. :smalltongue:

[TuggyNE]

That, and repairing ability damage is fairly easy (depending on how much you have, of course) in 3.5. A dedicated caster with that system (probably a skill-focused character) would probably grab bind vestige and bind Naberius, or get an item that cast restoration a few times per day. *shrug* Again, not perfect. :smalltongue:

I consider those to be "cheesing your way out of the consequences", for the record.

[zlefin]

the simple way is clear: don't use an existing variable.
Make a spellcasting check that you have to pass.
Don't use caster level, don't use skill level, don't use class level, don't use ANY existing variables, most/all existing variables have ways to mess with them.
By making an entirely new variable, there won't be any boosters for it other than what you make, so you can monitor the supply of them much better, or preferably just not have any.

[Jeff the Green]

Wow, that's a tremendous cost. It doesn't seem like you'd want to cast spells unless there was basically no other option, or unless you cheesed your way out of the consequences somehow. I don't think I want to soft-ban casting, really.

Yeah, that's even worse than my Unprotected Casting skill (see extended sig). Mine only does damage to you (and disintegration, if you die from the damage), but that much Wisdom damage is crippling.

However... You could make it a special type of damage. It naturally goes away at 1 point/round and can't be healed any other way. This will let you throw around 1st level spells at will, and low-level spells easily, but you couldn't throw around high-level spells like you can now.

[TuggyNE]

the simple way is clear: don't use an existing variable.
Make a spellcasting check that you have to pass.
Don't use caster level, don't use skill level, don't use class level, don't use ANY existing variables, most/all existing variables have ways to mess with them.
By making an entirely new variable, there won't be any boosters for it other than what you make, so you can monitor the supply of them much better, or preferably just not have any.

So, Base Casting Bonus?


However... You could make it a special type of damage. It naturally goes away at 1 point/round and can't be healed any other way. This will let you throw around 1st level spells at will, and low-level spells easily, but you couldn't throw around high-level spells like you can now.

What would you call that, "ability singe"? :smallwink:

[eggynack]

So, Base Casting Bonus?

Nah, that's way too confusing. Instead, it should be called, "Base Arcane Bonus". Nothing confusing in that. Or maybe, you could call it, "Ability to Cast", or AC for short.

[Jeff the Green]

What would you call that, "ability singe"? :smallwink:

Ability parcooking.

[Deepbluediver]

I know I said giving everything SR was lame and inelegant, but upon thinking it over again, it might work — at least as long as it's unified with the idea that absolutely any spell has some sort of resistance to being cast. In other words, merely adding cursory SR to creatures fails because it's insufficient, but this idea sounds closer to what I'm thinking would work.

The biggest obstacle I had to resolve was finding numbers that worked, both for a level-based bonus and for DC you had to beat, since you had to get all three (bonus, SR spells, non-SR spells) to work at the same time and with similar optimization levels.

There was a lot of trial and revision, and it's still not perfect, but it's functional at the very least, and IMO more so than the RAW.


One of the earlier posts reminded me of a magic system that I was working on a couple months ago (I never finished it). The idea essentially made magic another subsystem available to every class, doing away with all of the full-casting classes.
....
To cast a spell, you also needed an Int and Cha bonus equal to the spell level. (An Int&Cha score of 16 to cast 3rd level spells, for example). Getting 9th level spells is difficult (although not impossible) for humans, even with good starting scores.

I never liked the "You must be this tall to ride cast" method of doing things, because it made character development feel strictly limited, and it incentivized characters to be very SAD.

For my magic system, I based bonus spells off of Intellect, and the check/DC off of Wisdom for all casters. I figured that Int and Wis could be the equivalent of Constitution and Strength for casting characters. (there was also a feat to let you use Charisma in place of Wisdom, serving a similar purpose to Weapon Finesse and Dexterity).

Some one complained that this made casters feel to much the same, which I found odd, but I thought it was much less restrictive than the model that had all wizards stacking Int and all sorcerers stack Cha. Under my system, you could be a wizard with Intellect 6, if that's the way you wanted to play it, so that the choice of any stat was almost entirely up to the player (which is how it should be ideally, IMO).
Technically I guess you need a Wisdom score of at least 7 to cast 9th level spells, but even that is not absolute, since I was rewriting spells to include critical successes and failures (with auto-success on a natural 20).


the simple way is clear: don't use an existing variable.
Make a spellcasting check that you have to pass.
Don't use caster level, don't use skill level, don't use class level, don't use ANY existing variables, most/all existing variables have ways to mess with them.
By making an entirely new variable, there won't be any boosters for it other than what you make, so you can monitor the supply of them much better, or preferably just not have any.

This is like what I was saying. At it's simplest, the kind of thing you can slot into a game 5 minutes before it starts, is just a straight casting stat check against 10+spell level. You would have to optimize so heavily to auto-pass checks (a +11 bonus at level one, +19 at level 17) that I suspect most DM's would throw the book at you halfway through your character introduction.

The problem is that particular methods either leaves SR out in the cold or requiring a second check, but I think it's the kind of starting point people should think about when designing a system
That's essentially what mine is, just with a lot more modifiers and more complicated numbers; it's less easy to learn but more robust and complete in the number of situations it can handle.


Nah, that's way too confusing. Instead, it should be called, "Base Arcane Bonus". Nothing confusing in that. Or maybe, you could call it, "Ability to Cast", or AC for short.

Since I yanked "Spellcraft" off the skills list, I just called it "Base Spellcraft Bonus", or BSB. This was intentionally similar to BAB, because that was what designed it to be like.

If you felt that had to much potential for confusion, then we can hit up the thesaurus and rename it something like - PSB, or "Principle Spellcraft Bonus"


Edit: If you want to use a method that is more like skills, with various schools of magic that you can put points into, it might be workable, but you would still want to separate it from the existing skill system. Otherwise it ends up being a tax on caster's skill points and vulnerable to existing optimization methods.

So maybe you would get a number of spell-points per level, depending on class, and you could put them into things like Necromancy, Conjuration, etc.
And almost any system can be combined with Will or Fortitude saves if you like the versions that cause casters to risk health, sanity or other things in exchange for casting. If you can't get it down to just one unified mechanic, we can combine ideas to help with balance, power, and drawbacks.

I've talked a lot about what I did, but if you give me some further feedback on what you like or don't like, I'll try to help you come up with a new or modified system that is satisfactory.